... has been submitted for AERE 2011. Check it out below the fold (it's awesome). Let's hope that I can make it onto, at least, the poster session. I decided against submitting something for the 2011 NAAFE meeting. There is no way the state will fund a conference trip to Hawaii in this budget atmosphere (and the Whitehead Endowment won't cover it either). Rats.
Paper Title: WILLINGNESS-TO-PAY FOR OYSTER CONSUMPTION RISK REDUCTIONS BEFORE AND AFTER THE DEEPWATER HORIZON OIL SPILL
Select Session Type: Contributed Session, Poster Session
Abstract: In this paper we estimate the value of post-harvest processed oysters with the referendum format of the contingent valuation method. Using data from internet surveys of oyster consumers that were conducted before and after the Deepwater Horizon oil spill we consider two samples of respondents: those at risk-of-death from consuming raw oysters and those not at risk. With these two distinct samples we are able to measure willingness-to-pay for those directly affected by the risk reduction policy and those oyster consumers who are not at risk and who would not be affected. With a few assumptions, we are able to develop value of statistical life (VSL) estimates for both groups. With these data we are considering a number of other issues including the importance of scope effects, respondent uncertainty, preference heterogeneity, the effects of various information treatments and the appropriate integration of revealed and stated behavior data with the willingness-to-pay
data.
The literature on the VSL is vast. Researchers use the hedonic price method, the averting behavior method and the contingent valuation method. One concern with this literature is its policy application. Hedonic wage studies develop VSL estimates in the context of job-related risk. Averting behavior studies develop estimates in the context of traffic safety and consumer products. Application of job and traffic safety-based VSL estimates to risk reduction policies in other contexts may lead to inefficient policy. In the case of food safety, the relevant VSL may be less than or greater than the $4.3 million to $6.1 million consensus estimates developed from averting behavior and hedonic wage literature reviews. Differences may arise due to socioeconomic characteristics of the target population (e.g., age, income, health) and characteristics of the activity (traffic accidents vs. food consumption). The contingent valuation method can be used to develop VSL estimates for any type of risk, including food safety. In the food safety literature there are relatively few published estimates of the willingness-to-pay for risk reductions and no estimates of VSL that we are aware.
The policy being valued is for post-harvest processing of oysters to combat V. vulnificus, a bacterium found naturally in coastal waters along the Gulf, Atlantic, and Pacific coasts. V. vulnificus can be transmitted to humans through the consumption of raw shellfish harvested from waters containing the organism with Gulf of Mexico oysters of primary concern. Oyster consumers with compromised immune systems are at a greater risk for contracting V. vulnificus infections from oyster consumption. Life threatening illness in at risk individuals from consuming oysters arises primarily if the oysters are consumed raw or in an undercooked state. Between 1995 and 2009, the FDA recorded 510 V. vulnificus-related serious illnesses. Over 98 percent of these were associated with consuming raw oysters, of which 252 resulted in death. At an annual rate, approximately 30 consumers become seriously ill from consuming raw Gulf oysters, of which 15 to 20 die.
We developed an internet-based survey of oyster consumers (aged 18 and over), sampled Florida, Alabama, Georgia, Mississippi, Louisiana, Texas, and California. The sample was drawn from a panel of online respondents maintained by Online Survey Solutions, Inc. and the survey was administered in April 2010 and completed as the Deepwater Horizon oil spill began to unfold. Respondents were asked questions about awareness and perceptions of oyster consumption health risk, revealed and stated preference questions about oyster consumption, and a contingent valuation scenario. The sample size is almost 2000 with 1849 available for the CVM analysis after cases with significant item nonresponse are dropped.
The willingness-to-pay question is in the referendum format: "Suppose that in order to minimize the risks from eating raw oysters, the U.S. Food and Drug Administration (FDA) proposes a federal law to ensure that all oysters are post-harvest processed (PHP) before going to market. It is believed that this will reduce the average annual number of deaths in the U.S. from eating raw oysters from the current 16 to 20 people to [X] people. However, because of the additional costs incurred by oyster producers to process their product, the program will result in an increase in the price of an average oyster meal for all consumers. Imagine that you have the opportunity to vote on this proposed law. If more than 50% of those voting vote for the federal law, the FDA would put it into practice. If you could vote today and you knew that the price of your average oyster meal would go up by [bid] but the price of all other food would stay the same, would you vote for or against the propo sed law?" There are three randomly assigned versions of the annual deaths [X], 1 to 5, 6 to 10 and 11 to 15, and four randomly assigned bid levels, $1, $3, $5 and $7. Respondents were given three choice options, for, against and donâ?Tt know (would not vote), and a follow-up certainty question.
Preliminary empirical results are promising. First, the percentage of "for" votes falls significantly with the bid amount and increases significantly with the scope of the policy (the number of deaths avoided). A preliminary estimate of the VSL (assuming 6-12 oysters per meal and the annual Gulf oyster harvest) yields a plausible range of $2.4 million to $4.8 million. There are two surprising results. The first is that the policy is an inferior good. However, in a random parameter model we find that the income effect covers both the normal and inferior good range. The second is that stated preference oyster consumption falls with the risk reduction, apparently due to concerns about the effect of processing on oyster taste and texture. The next step in the analysis is to split the sample into those who are at-risk from eating raw oysters and those who are not at-risk and conduct analyses for willingness-to-pay for the public/private good for those at-risk and the public good for those not at-risk. This comparison should yield insights about the extent of altruism associated with this policy. We also plan to explore joint estimation of the stated preference consumption and willingness-to-pay data to further refine the role of altruism.
In December we resurveyed a portion of the respondents to the first survey. Our scenarios differed, reflecting issues raised by the Gulf oil spill about seafood safety, but the added CVM scenarios on traceability and aquaculture might yield additional insights concerning the issues raised above. As we have just received the data, preliminary results are not yet available.
Keyword(s): contingent valuation, value of statistical life, oyster consumption
Field(s): Environmental valuation + CBA: Contingent valuation, Resources and Ecosystem Studies: Fisheries, Resources and Ecosystem Studies: Health and environment
Your Role: Presenter
Addittional Authors:
Huth, William, University of West Florida (Co-Author)
Martin, Greg, Northern Kentucky University (Co-Author)
Morgan, Ash, Appalachian State University (Co-Author)
Sjolander, Richard, University of West Florida (Co-Author)
Select Session Type: Contributed Session, Poster Session
Abstract: In this paper we estimate the value of post-harvest processed oysters with the referendum format of the contingent valuation method. Using data from internet surveys of oyster consumers that were conducted before and after the Deepwater Horizon oil spill we consider two samples of respondents: those at risk-of-death from consuming raw oysters and those not at risk. With these two distinct samples we are able to measure willingness-to-pay for those directly affected by the risk reduction policy and those oyster consumers who are not at risk and who would not be affected. With a few assumptions, we are able to develop value of statistical life (VSL) estimates for both groups. With these data we are considering a number of other issues including the importance of scope effects, respondent uncertainty, preference heterogeneity, the effects of various information treatments and the appropriate integration of revealed and stated behavior data with the willingness-to-pay
data.
The literature on the VSL is vast. Researchers use the hedonic price method, the averting behavior method and the contingent valuation method. One concern with this literature is its policy application. Hedonic wage studies develop VSL estimates in the context of job-related risk. Averting behavior studies develop estimates in the context of traffic safety and consumer products. Application of job and traffic safety-based VSL estimates to risk reduction policies in other contexts may lead to inefficient policy. In the case of food safety, the relevant VSL may be less than or greater than the $4.3 million to $6.1 million consensus estimates developed from averting behavior and hedonic wage literature reviews. Differences may arise due to socioeconomic characteristics of the target population (e.g., age, income, health) and characteristics of the activity (traffic accidents vs. food consumption). The contingent valuation method can be used to develop VSL estimates for any type of risk, including food safety. In the food safety literature there are relatively few published estimates of the willingness-to-pay for risk reductions and no estimates of VSL that we are aware.
The policy being valued is for post-harvest processing of oysters to combat V. vulnificus, a bacterium found naturally in coastal waters along the Gulf, Atlantic, and Pacific coasts. V. vulnificus can be transmitted to humans through the consumption of raw shellfish harvested from waters containing the organism with Gulf of Mexico oysters of primary concern. Oyster consumers with compromised immune systems are at a greater risk for contracting V. vulnificus infections from oyster consumption. Life threatening illness in at risk individuals from consuming oysters arises primarily if the oysters are consumed raw or in an undercooked state. Between 1995 and 2009, the FDA recorded 510 V. vulnificus-related serious illnesses. Over 98 percent of these were associated with consuming raw oysters, of which 252 resulted in death. At an annual rate, approximately 30 consumers become seriously ill from consuming raw Gulf oysters, of which 15 to 20 die.
We developed an internet-based survey of oyster consumers (aged 18 and over), sampled Florida, Alabama, Georgia, Mississippi, Louisiana, Texas, and California. The sample was drawn from a panel of online respondents maintained by Online Survey Solutions, Inc. and the survey was administered in April 2010 and completed as the Deepwater Horizon oil spill began to unfold. Respondents were asked questions about awareness and perceptions of oyster consumption health risk, revealed and stated preference questions about oyster consumption, and a contingent valuation scenario. The sample size is almost 2000 with 1849 available for the CVM analysis after cases with significant item nonresponse are dropped.
The willingness-to-pay question is in the referendum format: "Suppose that in order to minimize the risks from eating raw oysters, the U.S. Food and Drug Administration (FDA) proposes a federal law to ensure that all oysters are post-harvest processed (PHP) before going to market. It is believed that this will reduce the average annual number of deaths in the U.S. from eating raw oysters from the current 16 to 20 people to [X] people. However, because of the additional costs incurred by oyster producers to process their product, the program will result in an increase in the price of an average oyster meal for all consumers. Imagine that you have the opportunity to vote on this proposed law. If more than 50% of those voting vote for the federal law, the FDA would put it into practice. If you could vote today and you knew that the price of your average oyster meal would go up by [bid] but the price of all other food would stay the same, would you vote for or against the propo sed law?" There are three randomly assigned versions of the annual deaths [X], 1 to 5, 6 to 10 and 11 to 15, and four randomly assigned bid levels, $1, $3, $5 and $7. Respondents were given three choice options, for, against and donâ?Tt know (would not vote), and a follow-up certainty question.
Preliminary empirical results are promising. First, the percentage of "for" votes falls significantly with the bid amount and increases significantly with the scope of the policy (the number of deaths avoided). A preliminary estimate of the VSL (assuming 6-12 oysters per meal and the annual Gulf oyster harvest) yields a plausible range of $2.4 million to $4.8 million. There are two surprising results. The first is that the policy is an inferior good. However, in a random parameter model we find that the income effect covers both the normal and inferior good range. The second is that stated preference oyster consumption falls with the risk reduction, apparently due to concerns about the effect of processing on oyster taste and texture. The next step in the analysis is to split the sample into those who are at-risk from eating raw oysters and those who are not at-risk and conduct analyses for willingness-to-pay for the public/private good for those at-risk and the public good for those not at-risk. This comparison should yield insights about the extent of altruism associated with this policy. We also plan to explore joint estimation of the stated preference consumption and willingness-to-pay data to further refine the role of altruism.
In December we resurveyed a portion of the respondents to the first survey. Our scenarios differed, reflecting issues raised by the Gulf oil spill about seafood safety, but the added CVM scenarios on traceability and aquaculture might yield additional insights concerning the issues raised above. As we have just received the data, preliminary results are not yet available.
Keyword(s): contingent valuation, value of statistical life, oyster consumption
Field(s): Environmental valuation + CBA: Contingent valuation, Resources and Ecosystem Studies: Fisheries, Resources and Ecosystem Studies: Health and environment
Your Role: Presenter
Addittional Authors:
Huth, William, University of West Florida (Co-Author)
Martin, Greg, Northern Kentucky University (Co-Author)
Morgan, Ash, Appalachian State University (Co-Author)
Sjolander, Richard, University of West Florida (Co-Author)
